lable at ScienceDirect

Journal of Stored Products Research 68 (2016) 63e72

Contents lists avai

Journal of Stored Products Research

journal homepage: www.elsevier .com/locate/ jspr

Application of radio frequency pasteurization to corn (Zea mays L.):Heating uniformity improvement and quality stability evaluation

Ajuan Zheng a, Bo Zhang a, Liyang Zhou a, Shaojin Wang a, b, *

a Northwest A&F University, College of Mechanical and Electronic Engineering, Yangling, Shaanxi 712100, Chinab Department of Biological Systems Engineering, Washington State University, 213 L.J. Smith Hall, Pullman, WA 99164-6120, USA

a r t i c l e i n f o

Article history:Received 24 April 2016Accepted 25 April 2016

Keywords:CornHeating uniformityPasteurizationQualityRadio frequency

* Corresponding author. Northwest A&F UniversityElectronic Engineering, Yangling, Shaanxi 712100, Ch

E-mail address: shaojinwang@nwsuaf.edu.cn (S. W

http://dx.doi.org/10.1016/j.jspr.2016.04.0070022-474X/© 2016 Elsevier Ltd. All rights reserved.

a b s t r a c t

Mildews caused grain losses and serious outbreaks have been becoming increasing concerns in domesticand international corn processing industry. This study intended to explore the possibility of using radiofrequency (RF) heating as an effective treatment to eliminate the mold contamination and reduce thedamage to corn quality. A pilot-scale, 27 MHz, 6 kW RF unit was used to study the heating uniformity incorn samples with five moisture contents (MC) and using three plastic material containers, and develop atreatment protocol for a corn sample with the MC of 15.0% w.b. and evaluate quality attributes andstorage stability of treated samples. The results showed that only 7.5 min was needed to raise the centraltemperature of 3.0 kg corn samples from 25 �C to 70 �C using the RF energy, but 749 min for samples toreach 68.6 �C using hot air at 70 �C. The RF heating uniformity was improved by adding forced hot air,moving samples on the conveyor, and mixing during the treatment. An effective RF treatment protocolwas finally developed to combine 0.8 kW RF power with a forced hot air at 70 �C, conveyor movement at6.6 m/h, two mixings, and holding at 70 �C hot air for 14 min, followed by forced room air coolingthrough thin-layer (2 cm) samples. Corn quality was not affected by RF treatments since quality pa-rameters of RF treated samples were better than or similar to those of untreated controls after theaccelerated shelf life test. RF treatments may hold great potential as a pasteurization method to controlmolds in corns without causing a substantial loss of product quality.

© 2016 Elsevier Ltd. All rights reserved.

1. Introduction

Corn or maize (Zea mays L.) is a plant of American origin andconsumed worldwide for various purposes (Velu et al., 2006).Approximately 75% of corn productions are used for feeding ani-mals and humans, and 25% are processed into various value-addedproducts (FAOSTAT, 2013). Corn kernels are rich and inexpensivesources of starch (62%) and protein (7.8%) (Singh et al., 2003), andare abundant sources of mineral elements (Gu et al., 2015), such asGu, Fe, Mn, K, and so on. Due to high nutrition values, internationalexport quantity and value of corn had been increased from 2009 to2011 in the international market (FAOSTAT, 2013). China is thesecond largest corn producing country with 218 million Mt cornyield, about 11.5% of the total world production (FAOSTAT, 2013).Therefore, the corn storage quality and production losses due to

, College of Mechanical andina.ang).

mold contamination are becoming major concerns of the cerealindustry.

About 25% of grains worldwide are infected by mold and itsmycotoxin each year, of which 2% loses its nutritional andeconomical values owing to serious infections (FAO, 2007). Asper-gillus parasiticus is one of the most common molds infecting cornduring storage (Reddy et al., 2009). Corn contaminations byAspergillus parasiticus, and consequently presences of the aflatoxinsare unavoidable during storage (Liu et al., 2006), which is a threat tohealth of human and animal (Massey et al., 1995; Bankole et al.,2010). However, it is difficult to remove aflatoxins from corns dueto their stabilized molecular structures. Eliminating Aspergillusparasiticus before the aflatoxins are produced must be the actualgoal rather than removing aflatoxins. It is, therefore, of a great in-terest to develop a postharvest processing method to reduce orcompletely eliminate fungus in corns before aflatoxins are pro-duced during storage.

There are a large number of suggested potential methods forsuppressing fungal growth and reducing mycotoxin formation.Except chemical treatments with ethylene (Gunterus et al., 2007),

A. Zheng et al. / Journal of Stored Products Research 68 (2016) 63e7264

physical methods are commonly applied, such as plasma (Hertwiget al., 2015), irradiation (UV, gamma rays) (Kanapitsas et al., 2015),microwave (Mendez-Albores et al., 2014) and modified atmo-spheres (Skandamis and Nychas, 2001). Although these can reducesurface fungal contaminations in some extent, each of thesephysical methods has limitations in terms of low efficiency, highcost and negative effects on product quality. Thus, it is desirable todevelop an effective, practical, and environmentally-friendlyphysical method for corn pasteurization.

Novel thermal treatments using radio frequency (RF) energyhold potential for pathogen control in agricultural commodities. RFenergy can directly interact with agricultural commodities torapidly raise the temperature of a whole treated sample volume inindustrial systems (Tang et al., 2000). A major advantage of RFenergy is deeper penetration depth and better heating uniformityin bulk materials (Wang et al., 2003) compared with microwaveheating. Thermal effects of RF treatments are found to be effectivefor controlling Salmonella in in-shell almonds (Gao et al., 2011) andinactivating molds, Salmonella and Escherichia coli on bread, blackand red pepper spices, respectively, with acceptable productquality (Liu et al., 2011; Kim et al., 2012). But there have been noreported studies on RF control of Aspergillus parasiticus in corns.

Heating uniformity is an important consideration in developingand scaling-up RF treatment protocols. Factors resulting in non-uniform RF heating include non-uniform electromagnetic fielddistribution and different dielectric properties between samplesand surrounding medium mainly caused by the MC of variousproducts (Wang et al., 2007a). Many practical methods are used toimprove the uniformity of RF heating in agricultural products, suchas adding forced hot air for sample surface heating, samplemovement, rotation or mixing of samples during RF heating andimmersing products into water for fresh fruits (Birla et al., 2004;Tiwari et al., 2008; Sosa-Morales et al., 2009; Zhou et al., 2015;Zhou and Wang, 2016; Ling et al., 2016). Research is needed toimprove the RF heating uniformity in developing corn postharvesttreatment protocols for pasteurization while maintaining theproduct quality.

Objectives of this study were (1) to analyze heating rates in cornunder two heating conditions (hot air and RF heating), and developan effective cooling method after RF heating, (2) to compare thetemperature distribution and the heating uniformity index in cornafter RF heating with additional forced hot air surface heating,sample movement, and mixing, (3) to determine the heating uni-formity of corn samples with five different MC and three plastic-type containers, and (4) to evaluate the effect of RF pasteurizationon the quality attributes (MC, water activity, ash, color, starch,protein, fat, fatty acid) of corns during an accelerated storage.

2. Materials and methods

2.1. Material and sample preparation

Newly harvested yellow corn (Zea mays L.) used in this researchwas purchased from a local farmer's market in Yangling, Shaanxi,China. The average initial MC of cornwas 9.1 ± 0.2%wet basis (w.b.).To explore the RF heating uniformity in corns with differentmoisture levels, other samples with MC of 12.0%, 15.0%, 17.9%, and20.9% w.b. were prepared for the experiment. The initial sampleswere conditioned by direct addition of predetermined amount ofdistilled water to obtain the targeted MC (Wang et al., 2015). Thepreconditioned samples were shaken by hands for 10min. Then thesamples were sealed in hermetic plastic bags and stored at 4 ± 1 �Cfor more than 7 days in a refrigerator (BD/BC-297KMQ, MideaRefrigeration Division, Hefei, China) for equilibrium. During thestorage, the bags were shaken 6 times per day. Before each test,

samples were placed in an incubator (BSC-150, Boxun Industry &Commerce Co., Ltd, Shanghai, China) to equilibrate for 12 h at25 ± 0.5 �C.

2.2. Hot air-assisted RF heating system

A 6 kW, 27.12 MHz pilot-scale free-running oscillator RF system(SO6B, Strayfield International, Wokingham, U.K.) with a hot airsystem (6 kW) was used for this study. A detailed description of theRF unit, the hot air and conveyor systems can be found in Wanget al. (2010) and Zhou et al. (2015). The top electrode(40 cm � 83 cm) was moved to obtain the required electrode gap,thus regulating the outputted RF power. To simulate continuousprocesses, the corn samples between electrodesweremoved on theconveyor belt started from inlet side to outlet side of the RF system.The sample surface heating was provided using the hot air duringRF heating and holding without the RF power. The hot air speedwas setup to be 1.6 m/s inside the RF cavity and measured at 2 cmabove the bottom electrode by an anemometer (DT-8880, ChinaEverbest Machinery Industry Co., Ltd, Shenzhen, China).

2.3. Determining heating temperature and holding time

For developing effective pasteurization processes, the RF treat-ment protocol parameters should be designed based on the ther-mal death kinetics of target molds (Buzrul and Alpas, 2007). Jinet al. (2011) studied the inactivation kinetic model of Aspergillusparasiticus in moldy corn by microwave processing, and reportedthat 6-log reductions of Aspergillus parasiticus spores in mL sus-pensions were obtained at 50, 55, 60, 65, 70 and 75 �C for 19.6, 15.6,15.2, 13.6, 13.2 and 13.1 min, respectively. Fang et al. (2011) inves-tigated the effect of microwave radiation on biochemical charac-teristics and mortality of Aspergillus parasiticus compared to effectsof conventional heating treatment, and the research results showedthat effects of microwave treatments on inactivating Aspergillusparasiticus were also determined around 70 �C. Taking intoconsideration of the required inactivation of Aspergillus parasiticus,non-uniformity of RF heating, the corn quality and treatment effi-ciency, the target sample temperature of 70 �C was selected todevelop the RF treatment protocol.

2.4. Material selection of surrounding containers

To study the effect of different surrounding containers on the RFheating uniformity of corns, three kinds of materials (poly-etherimide, polystyrene, and polypropylene) were selected basedon the reported improvement of RF heating uniformity (Jiao et al.,2014; Huang et al., 2016). The containers had the same inner di-mensions (300 mm � 220 mm � 60 mm) (Fig. 1) with the 5 mmthickness, which consisted of perforated screens on the side andbottomwalls to allow hot and room air to pass through the samplesfor surface heating and cooling. Effects of the three-material con-tainers with different dielectric properties on RF heating uniformitywere compared.

2.5. Determining electrode gap and conveyor belt speed

A corn sample of 3.0 kg with MC of 15.0% w.b. was used for fullloads in a polypropylene container (300 mm � 220 mm � 60 mm)(Fig. 1). This MC level was selected since its water activity wasclosed to that in the growth boundary of Aspergillus parasiticus atroom temperature. To determine an appropriate electrode gap forRF treatments, the polypropylene container was placed on thestationary conveyor belt between the two electrodes to obtain ageneral relationship between electrode gap and electric current (I,

Fig. 1. Rectangular plastic container with 14 locations sample temperature measure-ments (all dimensions are in mm).

A. Zheng et al. / Journal of Stored Products Research 68 (2016) 63e72 65

A). The range of the electrode gap was from 10.0 cm to 19.0 cmwitha 1.0 cm interval. A detail description of measurement could befound in Zhou et al. (2015). All tests were repeated three times.Based on the measured electric current, three electrode gaps of10.0 cm, 11.0 cm and 12.0 cm were selected for furthertemperature-time history experiments in RF treated corns based onno arcing and fast heating rates. To study the RF heating uniformityand treatment protocol, the best one was determined from thethree selected electrode gaps. The time needed to heat the corns inthe container center from 25 �C to 70 �C was recorded, and thesample temperature was measured at the geometric center of thecontainer using a six-channel fiber-optic temperature sensor sys-tem (HQ-FTS-D120, Heqi Technologies Inc., Xian, China) with anaccuracy of ±0.5 �C. The final electrode gap for the samples withMCof 15.0% w.b. was fixed based on the target heating rate (4e6 �C/min) with three replicates.

For the corn samples with different MC, the different electrodegaps were selected to achieve the similar effect of heating rates onthe RF heating uniformity. To determine the time-gap relationshipin RF treated corn samples with other four MC in the polypropylenecontainer, three electrode gaps of 10.0, 10.5 and 11.0 cm werechosen for corn samples with MC of 9.1% and 12.0% w.b., respec-tively. Meanwhile, three electrode gaps of 11.0, 11.5 and 12.0 cmwere selected for corn samples with MC of 17.9% and 20.9% w.b.,respectively. The most suitable gap was obtained based on theclosest heating rate of corn samples at 15.0% w.b.. Then theconveyor belt speed was calculated with dividing the electrodelength by the resulted heating time.

2.6. Comparisons of temperature profiles of corns between hot airand RF heating

The most appropriate gap determined in section 2.5 was usedfor temperature profile comparisons, heating uniformity improve-ment and protocol development. The polypropylene container with3.0 kg samples at 15.0% w.b. was placed on the bottom electrode forhot air and RF heating. The temperature at the geometric center ofsamples was recorded every 60 s by the fiber-optic temperaturesensor system when subjected to forced hot air heating at 70 �Cwith speed of 1.6 m/s and RF treatments with the electrode gap of11.0 cm. The measurement was stopped when the sample tem-perature reached 70 �C for RF heating or the temperature increasewas less than 0.5 �C within 30 min for hot air heating. Ambientroom temperature (25 �C) was used as the initial sample temper-ature, and each test was repeated three times.

2.7. Determination of cooling methods

Appropriate cooling method is of great importance to avoidquality degradation and improve processing efficiency for devel-oping RF treatment protocols (Wang et al., 2010; Gao et al., 2011).The polypropylene container filled with 3.0 kg corn samples withMC of 15.0% w.b. heated to 70 �C by hot air was used to develop thesuitable cooling method. Corn samples with 6, 3 and 2 cm depthsheld in the polypropylene container were subjected to naturalambient air and forced room air cooling. The forced room air wasprovided by an electric fan (FT20-10A, Guangdong Midea Envi-ronment Appliances Manufacture Co., Ltd., Zhongshan, China). Themeasured air speeds on the sample surface were about 0.2 and3.5 m/s for the natural and forced air cooling, respectively. Thetemperature in the center position of samples was recorded every60 s until the sample temperature dropped to 30 �C. The bestcooling method was further used to determine the temperature-time histories of corns and develop RF treatment protocols.

2.8. Heating uniformity tests

2.8.1. Comparison of RF heating uniformity under differentoperational conditions

The heating uniformity is important for developing effective RFtreatments and depends on practical treatment conditions, such aswith or without forced hot air, with or without movement on theconveyor belt, and with or without mixing (Jiao et al., 2012; Zhouet al., 2015; Zhou and Wang, 2016). Tiwari et al. (2011) studiedthe effect of the sample position on the RF heating uniformity bythe computer simulation, and the results show that RF energydistribution is improved and the required heating uniformity isachieved when samples are placed exactly in the middle of the twoRF electrodes. The polypropylene container filled with 3.0 kg cornsamples with MC of 15.0% w.b. was placed in the middle of twoelectrode gaps and heated in the RF system to compare the sampletemperature distribution under seven conditions: (1) RF heatingonly, (2) RF heating with hot air assisted, (3) RF heating withconveyor belt movement, (4) RF heatingwith a singlemixing, (5) RFheating with two mixings, (6) RF heating with forced hot air of70 �C assisted, conveyor belt movement and a single mixing, and(7) RF heating with hot air, conveyor belt movement, two mixingsand holding in 70 �C hot air for 14 min. The container movementstarted from the right edge (in-feed side) to the left edge (out-feedside) of the electrode. The forced hot air at 70 �C was providedthrough the air distribution box at the bottom electrode. One andtwomixings were conducted outside the RF cavity at the interval ofhalf and one third of RF treatment time by hand in a large container(355 mm � 275 mm� 105 mm), and then the samples were placedback again into the treatment container and finally the RF cavity forthe remainder of the treatment time. The entiremixing process wasfinished within one minute. The sample at 15.0% w.b. with 3.0 kg inweight and 6 cm in depth in the polypropylene container weredivided into three layers (Fig. 1) and separated by two thin gauzes(with mesh opening of 1 mm) to easily map the sample surfacetemperatures with a thermal imaging camera (DM63, Zhejiang DaliTechnology Co., Ltd, Hangzhou, China) with an accuracy of ±2 �C.Before and immediately after RF treatments, the sample tempera-tures on the top, middle and bottom layers without mixing weremeasured by the thermal imaging camera. Each thermal image tookless than 2 s. Details on the infrared imaging system to measureproduct surface temperature after RF treatment can be found inWang et al. (2006a). The surface temperature data in the treatmentarea were obtained and used for estimating the average tempera-ture and standard deviation (SD). With mixing, the sample tem-peratures of middle and bottom layers at 14 selected positions were

A. Zheng et al. / Journal of Stored Products Research 68 (2016) 63e7266

obtained by twoType-T thermocouple thermometers (TMQSS-020-6, Omega Engineering Ltd., CT, USA). The 14 selected positions wereequally distributed at two depths of 4 and 6 cm (Fig. 1). The averageand SD values of the measured sample temperatures for eachreplicate were used for evaluating RF heating uniformity.

2.8.2. Validating the final RF protocol for corns with different MCand containers

Corn samples with the MC of 15.0% w.b. in the polyetherimideand polystyrene containers, respectively, were tested to examinethe effect of treatment conditions with conveyor belt movement,hot air, two mixings and holding in 70 �C hot air for 14 min on theRF heating uniformity. Corn samples with the MC of 9.1%, 12.0%,17.9%, and 20.9% w.b. in the polypropylene container, respectively,were also tested to determine the heating uniformity under thesame treatment conditions. Temperature measurements weremade according to the aforementioned methods. Each test wasrepeated three times. The average and SD values of the surfacetemperatures for each replicate were used for evaluating the RFheating uniformity.

2.8.3. Evaluating RF heating uniformityHeating uniformity is a key factor in developing successful

postharvest quarantine treatments using RF energy. Heating uni-formity index (l) has been successfully used for evaluating RFheating uniformity in different agricultural products, such as wal-nuts (Wang et al., 2007a), legumes (Wang et al., 2010; Jiao et al.,2012), and milled rice (Zhou et al., 2015). Heating uniformity in-dex is defined as the ratio of the rise in standard deviation ofsample temperatures to the rise in average sample temperaturesduring treatment and can be calculated by the following equation(Wang et al., 2008):

l ¼ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffis2 � s20

q

m� m0(1)

Where m0 and m are initial and final mean corn temperatures (�C), s0and s are initial and final standard deviations (�C) of corn tem-peratures over treatment time, respectively. The smaller valuesrepresent the better RF heating uniformity.

2.9. Treatment protocol development

The optimal RF treatment protocol could be developed withpreviously determined suitable electrode gap, heating uniformitytests and cooling methods. Based on the above tests, the electrodegap (11.0 cm) was used in the RF system together with conveyorbelt movement at the speed of 6.6 m/h, and additional hot airheating at 70 �C, and a twice mixings to heat 3.0 kg of corn sampleswith the MC of 15.0% w.b. and 6 cm sample thickness in the poly-propylene container. After the sample center temperature in thecontainer reached 70 �C, RF was turned off, and the corn sampleswere held in hot air at 70 �C for 14 min, followed by forced room aircooling with thin-layer (2 cm in depth) for 33 min. The untreatedsamples were considered as controls. Each treatment was repeatedthree times. The treated and untreated samples were sealed in bagsfor storage tests and quality evaluations.

2.10. Storage experiment and quality evaluations

Before and after RF treatments, the quality of controls and RFtreated corn samples with the MC of 15.0% w.b. in the poly-propylene container was evaluated immediately and after acceler-ated shelf life storage. Corn usually has the relatively low

temperature (10 �C) storage after harvest, and new corns would beavailable one year later (Wang and Zhang, 2011), so longer storagewas not considered in this study. Corn samples (600 g) were packedindividually in 10 bags and stored in the incubator set to 35 ± 0.5 �Cwith 68 ± 0.5% relative humidity (RH) for 17 d to simulate com-mercial storage at 10 �C for 1 year. The accelerated storage time at35 �C was calculated based on a Q10 value of 3.41 for nutrition loss(Taoukis et al., 1997), which was validated by real-time storageexperiments (Wang et al., 2006b). This accelerated tests was suc-cessfully used for simulating the real storage experiments in RFtreated products (Wang et al., 2002; Gao et al., 2010; Jiao et al.,2012; Hou et al., 2014; Ling et al., 2016; Zhou et al., 2015; Zhouand Wang, 2016). After 0, 4, 8, 12 and 17 d of storage, the cornsamples were withdrawn from each treatment for quality analysis.MC, water activity, ash, color, starch, protein, fat, and fatty acid werechosen to analyze the corn quality change.

A computer vision system (CVS) was used for measuring colorvalues of control and RF treated samples at 15.0% w.b. in thepolypropylene container. A detail description of the CVS unit can befound in Hou et al. (2014) and Zhou et al. (2015). Based on themeasurement procedures suggested by Ling et al. (2015) and toreduce the data variability, the corns were ground in a universalhigh-speed blender (FLB-100, Feilibo Food machinery co., Ltd.,Shanghai, China) and placed in a plastic Petri dish (8 cm diameter)for measurement. Color images of corn flour (20 g) surface pertreatment were captured and stored in the computer, thenanalyzed by Adobe Photoshop CS3 (Adobe Systems Inc., USA). Thecolor values: lightness (L), redness-greenness (þ or e a) andyellowness-blueness (þ or e b), which were obtained from Pho-toshop could be converted to CIE LAB (L*, a* and b*) values (Brionesand Aguilera, 2005). The L* value changes from 0 to 100 repre-sented dark to light. A higher positive a* and b* values indicatedmore redness and yellowness, respectively (Sandhu et al., 2007).

To evaluate the influence of RF heating on chemical componentsof corn samples with 15.0%w.b. in the polypropylene container, MC,water activity, starch content, protein content, fat content, ashcontent, and free fatty acid content were chosen for chemicalanalysis.

The MC of corns was determined according to the oven dryingmethod described by AOAC (2000a). About 5 g corn samples withless than 5 mm thickness placed in aluminum dishes were dried ina vacuum oven (DZX-6020B, Nanrong Laboratory Equipment Co.,Ltd., Shanghai, China) at 105 �C, and 103.1 kPa until a constantweight of samples was attained. The corn samples were placed in adesiccator with CaSO4 for cooling. Until the temperatures of cornsamples dropped to room temperature, their weight was recorded.The MC was estimated from initial and final weights of the cornsamples. About 5 g of corn samples were placed into a sample cupand then in an Aqua Lab water activity meter (Model 4TE, DecagonDevices, Inc., Pullman, WA, USA) to measure water activity underambient temperature (25 �C).

Starch is one of the most important compositions of corns and avaluable ingredient to the food industry, being widely used as athickener, gelling agent, bulking agent and water retention agent(Ubwa et al., 2012). Corn starch was determined by using theenzymatic hydrolysis method following the AOAC Standard (AOAC,2000b). The protein was detected using the Kjeldahl methodfollowing the AOAC method (AOAC, 2005). A nitrogen conversionfactor of 6.25 was used to calculate the protein content of the corn(Malumba et al., 2009). Fat content in corn was measured by theSoxhlet extraction method following the AOAC method (AOAC,2006). Total ash content was determined by weighing the resid-ual ash obtained by combustion in a Muffle furnace at 550 �C for4 h. The fatty acid content is frequently determined as a qualityindex during production storage, and was determined by the

Fig. 3. Temperature-time histories of the RF heated corn samples with 15.0% w.b. inthe center of polypropylene containers as function of the electrode gap.

A. Zheng et al. / Journal of Stored Products Research 68 (2016) 63e72 67

manual titration method following the National Standard of China(GB/T 20570-2015, 2015).

2.11. Statistical analysis

Mean values and standard deviations were calculated from thethree replicates for each treatment. Differences were estimated bythe analysis of variance (ANOVA) followed by Tukey's test andconsidered significantly at P � 0.05. All statistical analyses wereperformed using the statistical software SPSS 16.0 version (SPSSInc., Chicago, IL, USA).

3. Results and discussion

3.1. Electric current as influenced by the electrode gap

Fig. 2 shows the relationship between the electric current andelectrode gap when the polypropylene container filled with orwithout corn samples at 15.0% w.b. was placed on the RF bottomelectrode without movement and forced hot air. Electric currentdecreased with increasing electrode gap (10.0e19.0 cm). Therepeatable current resulted in small standard deviation. Specially,electric current decreased rapidly from 0.56 A to 0.40 A when theelectrode gap increased from 10 cm to 13 cm, and then decreasedslowly when the electrode gap changed from 13 cm to 19 cm. Threeelectrode gaps (10.0, 11.0, and 12.0 cm) were then chosen for thefurther tests.

3.2. Determination of electrode gap and conveyor belt speed

Fig. 3 shows the temperature at the geometric center of poly-propylene container during the RF heating using three differentelectrode gaps. The corresponding time required to raise the sam-ple temperature from 25 �C to 70 �C was about 4.52, 7.50 and10.31 min for electrode gaps of 10.0, 11.0, and 12.0 cm, respectively.The temperature-time histories also demonstrated that heatingrates decreased with increasing electrode gaps. The heating rateswere 9.96, 6.00 and 4.36 �C/min for electrode gaps of 10.0, 11.0, and12.0 cm, respectively.

Fig. 4 shows the time required to raise the temperature at thegeometric center of polypropylene container from 25 �C to 70 �C for3.0 kg corn samples with MC of 9.1%, 12.0%, 17.9%, and 20.9% w.b.using the selected three electrode gaps, respectively. The timeneeded for the temperature to increase from 25 �C to 70 �C was

Fig. 2. The relationship between the electrical current and electrode gap with apolypropylene container filled with corn samples at the MC of 15.0% w.b.

5.80, 7.38 and 9.55min for the samples withMC of 9.1% at electrodegaps of 10.0,10.5 and 11.0 cm, respectively. These required timewasreduced to 5.72, 7.17 and 8.57 min for samples with MC of 12.0%w.b., respectively. Correspondingly, these required time was 5.53,6.33 and 7.72min for the samples withMC of 17.9%w.b. at electrodegaps of 11.0, 11.5, and 12.0 cm, which were further reduced to 4.97,5.30 and 7.35 min for samples with MC of 20.9%, respectively.

The RF heating time decreased with increasing MC of cornsamples under same electrode gap. Shorter heating time corre-spond to the higher throughputs, but the RF heating uniformitycould be adversely affected by high heating rates. To obtain rela-tively high throughputs with acceptable heating uniformity in in-dustrial applications, the electrode gap of 10.5 cm was determinedto complete the RF heating test for corn samples with MC of 9.1%and 12.0%, respectively. An electrode gap of 11.0 cmwas consideredas suitable operational parameters for the corn samples with MC of15.0% w.b. during RF heating, together with 12.0 cm for the cornsamples with MC of 17.9% and 20.9% w.b. These optimal electrodegaps were used for further experiments.

Fig. 5 shows the comparison of the temperature-time histories ofthe RF heated corn samples in the center of polypropylene containerat their optimal electrode gaps with five MCs. The heating rates ofcorn samples with MC of 9.1% and 12.0% w.b. for electrode gap of10.5 cmwere 6.09 and 6.27 �C/min, respectively. The heating rate ofcorn samples with MC of 15.0% w.b. was 6.00 �C/min for electrodegap of 11.0 cm. About 5.83 and 6.12 �C/min were considered asoptimal heating rates for the corn samples with MC of 17.9% and20.9%w.b., respectively. The heating rate range is generally requiredby RF heating uniformity and treatment throughput and is also usedinmany RF treatment protocols (Wang et al., 2007a; Jiao et al., 2012;Pan et al., 2012; Hou et al., 2014; Zhou et al., 2015).

The speed of conveyor belt was estimated under the threeelectrode gaps and five MC as shown in Table 1 together with totalRF heating time (sample entirely moved through the top electrodelength direction). The speed of conveyor belt was determined bythe heating rates and the electrode length and around 6.5e6.9 m/hunder different MC. The total RF heating time and the interval timeof 2 mixings were around 9.83e10.43 min and 3.28e3.48 min,respectively. These values could be further used for pasteurizingcorns in large-scale experiments.

3.3. Heating and cooling profiles

A comparison of the temperature-time histories of corn samples

Fig. 4. Time-gap relationship of the RF heated corn in the center of polypropylene containers with the given four moisture contents to achieve the final temperature of 70 �C.

Fig. 5. Temperature-time histories of the RF heated corn samples with five moisturecontents in the center of polypropylene containers under their optimal electrode gaps.

Table 1Parameters of the optimal electrode gap, movement speed (vmovement), total RFheating time (ttotal) and interval time (tinterval) of 2 mixings of corn samples duringthe RF treatment with the given five moisture contents.

Moisture contents (MC, % w.b.)

9.1 12.0 15.0 17.9 20.9

Gap (cm) 10.5 10.5 11.0 12.0 12.0vconveyor (m/h) 6.7 6.9 6.6 6.5 6.8ttotal (min) 10.1 9.8 10.3 10.4 10.0tinterval (min) 3.4 3.3 3.4 3.5 3.3

A. Zheng et al. / Journal of Stored Products Research 68 (2016) 63e7268

at 15.0% w.b. in the center of polypropylene container is shown inFig. 6 using only hot air heating at 70 �C and only RF heating withthe electrode gap of 11.0 cm. The heating time indicated a differ-ence between RF heating and conventional hot air heating for3.0 kg samples. Specifically, it took about 7.50 min for the centertemperature of corn sample to reach 70 �C under the RF heating,but about 749 min to increase from 25 �C to 68 �C with forced hotair heating at 70 �C. The slow hot air heating could be caused by thepoor heat conduction. The large heating time difference betweenRF and hot air heating in this study is similar to the results observedwith legumes (Wang et al., 2010), almonds (Gao et al., 2010), coffeebeans (Pan et al., 2012), chestnuts (Hou et al., 2014), milled rice(Zhou et al., 2015), and pistachios (Ling et al., 2016).

Fig. 6. Typical temperature-time histories of corn samples with 15.0% w.b. in thepolypropylene container when subjected to hot air heating at 70 �C and RF heating(electrode gap ¼ 11.0 cm).

A. Zheng et al. / Journal of Stored Products Research 68 (2016) 63e72 69

Fig. 7 shows the temperature-time histories of corn sampleswith 15.0% w.b. in the polypropylene container center with threesample thicknesses of 6, 3, and 2 cm under natural or forced roomair cooling. The cooling time decreased with increasing air speedand decreasing sample thickness. For example, when using naturalroom air cooling, about 196, 107, and 69 min were needed for 6, 3,and 2 cm thick samples, respectively, to cool down from 70 �C to30 �C. However, the time was reduced to 68, 51 and 33 min,respectively, when subjected to a forced room air with speed of3.5 m/s. This cooling time would further decrease by reducing thesample thickness in industrial applications. Compared to thecooling time of three sample thicknesses, the thin layer (2 cmdepth) corn samples with the forced room air resulted in theshortest one, which was further used to develop the RF treatmentprotocol.

3.4. Heating uniformity evaluation

Table 2 provides a detailed comparison of the temperaturedistribution and uniformity index values in the top, middle, andbottom layers for corn samples at 15.0% w.b. in the polypropylenecontainer after RF heating under different operational conditions.The initial sample temperature distribution was relatively uniformbefore RF treatment. After heating treatment, the average samplesurface temperature in the middle layer was the highest, followedby the bottom and top layers. The relatively lower average tem-peratures of the samples in the top and bottom layers were prob-ably caused by heat loss to ambient air. Conveyor belt movement,hot air assistance and mixing reduced the standard deviation oftemperatures in each layer and improved heating uniformity overthe RF only treatment. Forced hot air could help to raise the averagetemperature for all the locations with the decreased standard de-viations of sample temperatures and the heating uniformity indexvalues. Mixing was the best method to improve heating uniformityas compared to hot air andmovement based on the reduced l value.In the industrial-scale RF treatments, mixing could be achieved bymechanical tumbling between two RF units (Wang et al., 2007a,b).With holding, the average sample temperatures in the three layerswere reduced by about 2 �C, but the heating uniformity was furtherimproved due to reduced standard deviations. Thesewere probablycaused by the decrease of temperature difference between hot andcold spots by heat conduction. It was clear that the heating uni-formity index was the smallest under the condition of RF heating,sample movement, 2 mixings and holding 14 min, and the average

Fig. 7. Cooling curves of corn samples with 15.0% w.b. in the polypropylene containercenter as a function of sample thickness with natural and forced room air cooling.

temperatures were close to the target 70 �C for meeting thepasteurization requirement. Therefore, including hot air, move-ment, mixing and holding in the design of an RF treatment protocolwould be adequate to obtain the required heating uniformity. Theheating uniformity indexes of corn samples at 15.0% w.b. in thepolypropylene container are similar to those found for chestnuts(Hou et al., 2014), almonds (Gao et al., 2010), and milled rice (Zhouet al., 2015), but slightly larger than those observed for coffee beans(Pan et al., 2012), and lentils (Jiao et al., 2012), and smaller thanthose obtained for walnuts (Wang et al., 2007a), and pistachios(Ling et al., 2016).

Table 3 shows the comparison of temperature and heatinguniformity index of the corn samples at 15.0% w.b. in the threedifferent plastic containers after RF treatment. The temperatures ofcorn samples in the three layers were a little higher than 70 �C, andthere was no significant difference (P > 0.05) among the three-typecontainers for the average temperatures and heating uniformityindexes of corn samples. The surface temperatures of corn samplesin the polyetherimide container were the highest, but the standarddeviation of the sample temperatures and heating uniformity indexvalues were the lowest. Similar results are observed by Jiao et al.(2014) and Huang et al. (2015), indicating that a container havingthe similar dielectric constant and a negligible dielectric loss factorto those of samples can obtain the better RF heating uniformity. Afurther experiment is needed to evaluate the effect of the differentmaterial containers with various thicknesses on the RF heatinguniformity.

Table 4 shows a comparison of surface temperatures and heat-ing uniformity index of corn samples with five MC in the poly-propylene container after RF treatment. The lower MC resulted inthe better heating uniformity. For all corn samples, the averagesurface temperature and standard deviation, and the heating uni-formity index of corns with the high MC were both higher thanthose in low MC. This result could be caused by the differentdielectric properties of corn samples and electrode gaps. The pre-vious research results indicated that the smaller dielectric proper-ties lead to better RF power uniformities in the samples and dryproducts should provide a better uniform RF heating because of thesmaller dielectric properties (Tiwari et al., 2011). Heating unifor-mity index of corn samples decreased as the electrode gap wasreduced from 12.0 to 10.5 cm, which well agreed with the results inRF heated wheat flour (Tiwari et al., 2011).

3.5. Quality of RF heated corns

Table 5 shows the MC and water activity of corn samples at15.0% w.b. in the polypropylene container before and after RFtreatment during accelerated shelf life storage. A significant dif-ference (P < 0.05) was observed between control and RF treatedsamples in the entire storage process. The RF treated corn sampleslost about 1.6% w.b. of MC compared with control samples at thestorage time of 0 d. The MC of corn decreased with the increasingstorage time, and the MC of control samples significantly decreased(P < 0.05) at 4th d, but there were no significant differences(P > 0.05) in subsequent storage time. For RF treated samples,however, the MC significantly decreased after 8 d storage (P < 0.05).The changes of MC in control corn samples during storage weresimilar to those of corns during storage for 180 d at 10 �C reportedby Zia-Ur-Rehman (2006).

Table 6 provides a detailed comparison of major quality pa-rameters (protein, fat, starch, ash, and fatty acid) of corn sampleswith 15.0% w.b. in the polypropylene container before and after RFtreatment during the 17 d accelerated shelf life storage. There wasno significant difference (P > 0.05) between control and RF treatedsamples for five selected quality attributes except for fatty acid. The

Table 2Comparison of the temperature and heating uniformity index (mean ± SD over 3 replicates) of the corn sample with 15.0% w.b. in the polypropylene container after RFtreatments using different operational conditions.

Layer RF RF þ hot air RF þ movement RFþ 1mixing RF þ 2mixings

RF þ hotair þ movement þ mixing

RF þ hot air þ movement þ 2 mixings þ holding14 min

Temperature (�C)Top 69.8 ± 8.4 70.1 ± 8.1 70.4 ± 7.0 70.9 ± 4.9 70.4 ± 3.7 72.1 ± 2.4 70.9 ± 2.0Middle 73.6 ± 6.4 73.5 ± 6.1 73.1 ± 5.0 72.5 ± 3.0 71.6 ± 2.5 73.2 ± 2.1 71.2 ± 1.8Bottom 70.3 ± 6.4 71.4 ± 6.0 70.8 ± 5.3 71.0 ± 3.1 70.7 ± 2.4 73.0 ± 2.0 70.2 ± 1.9Heating uniformity index (l)Top 0.172 ± 0.002 0.165 ± 0.003 0.143 ± 0.004 0.094 ± 0.003 0.066 ± 0.001 0.053 ± 0.002 0.050 ± 0.003Middle 0.121 ± 0.004 0.115 ± 0.005 0.097 ± 0.003 0.057 ± 0.003 0.046 ± 0.003 0.042 ± 0.006 0.040 ± 0.004Bottom 0.136 ± 0.008 0.122 ± 0.007 0.111 ± 0.002 0.063 ± 0.004 0.052 ± 0.007 0.048 ± 0.003 0.042 ± 0.002

Table 3Comparison of temperature and heating uniformity index (mean ± SD over 3 rep-licates) in corn samples with 15.0% w.b. after hot air assisted RF treatments withconveyor movement, 2 mixings and holding for 14 min using three different plasticcontainers.

Layers Container materials

Polypropylene Polystyrene Polyetherimide

Temperature (�C)Top 70.9 ± 2.0a 71.0 ± 2.0 71.4 ± 3.0Middle 71.2 ± 1.8 73.0 ± 1.7 73.2 ± 1.7Bottom 71.1 ± 1.9 71.2 ± 1.9 72.6 ± 1.8Heating uniformity index (l)Top 0.050 ± 0.003 0.045 ± 0.001 0.041 ± 0.002Middle 0.040 ± 0.004 0.039 ± 0.002 0.035 ± 0.003Bottom 0.042 ± 0.002 0.040 ± 0.002 0.037 ± 0.001

a No significant difference was observed both for container materials and layers(P > 0.05).

Table 4Comparison of corn samples in the polypropylene container temperature and heating uniformity index (mean ± SD over 3 replicates) after hot air assisted RF treatments withconveyor movement, 2 mixings and holding for 14 min with five moisture contents (MC) under their optimal electrode gaps.

Layers MC (% w.b.)

9.1 12.0 15.0 17.9 20.9

Temperature (�C)Top 70.0 ± 1.9 70.3 ± 2.0 70.9 ± 2.0 70.3 ± 3.0 70.8 ± 3.0Middle 70.4 ± 1.5 70.7 ± 1.7 71.2 ± 1.8 71.1 ± 2.5 71.0 ± 2.9Bottom 70.3 ± 1.4 70.5 ± 1.4 70.2 ± 1.9 70.9 ± 2.7 70.9 ± 2.7Heating uniformity index (l)Top 0.041 ± 0.003 0.048 ± 0.002 0.050 ± 0.003 0.056 ± 0.003 0.065 ± 0.001Middle 0.034 ± 0.005 0.036 ± 0.008 0.040 ± 0.004 0.044 ± 0.007 0.055 ± 0.006Bottom 0.039 ± 0.002 0.040 ± 0.006 0.042 ± 0.002 0.051 ± 0.010 0.061 ± 0.004

Table 5Moisture contents (MC) (% w.b.) and water activities (Aw) (mean ± SD over 3 replicates) of corn samples with 15.0% w.b. in the polypropylene container before and after radiofrequency (RF) treatment.

Treatment Storage time (days)b

0 4 8 12 17

MC (% w.b.)Control 15.0 ± 0.2aAa 14.4 ± 0.1aB 14.2 ± 0.2aB 14.1 ± 0.1aB 14.0 ± 0.1aBRF 13.4 ± 0.1bA 13.1 ± 0.1bA 12.9 ± 0.2bB 12.7 ± 0.1bB 12.7 ± 0.1bBAw

Control 0.791 ± 0.006aA 0.780 ± 0.005aA 0.778 ± 0.009aB 0.767 ± 0.010aB 0.757 ± 0.008aBRF 0.731 ± 0.007bA 0.715 ± 0.009bA 0.693 ± 0.013bB 0.685 ± 0.010bB 0.667 ± 0.009bB

a Different lower and upper case letters indicate that means are significantly different at P ¼ 0.05 among treatments and storage time, respectively.b 17 days at 35 �C, 68% RH to simulate 1-year storage at 10 �C.

A. Zheng et al. / Journal of Stored Products Research 68 (2016) 63e7270

protein and starch values fluctuated in a narrow range withincreasing storage time both for control and RF treated corn sam-ples. Fat and ash contents decreased but fatty acids increased withincreasing storage time. Similar results were reported in RF treated

milled rice by Zhou et al. (2015), and Zhou andWang (2016). Duringstorage time, the decreased protein content is caused by reducingcontent of free amino nitrogen (Onigbinde and Akinyele, 1988). Thedecreased fat ash and contents in grains are also obtained by Zhouand Wang (2016) and Reed et al. (2007).

Table 6 shows that fatty acid contents significantly increasedwith increasing storage time (P < 0.05). There was no significantdifference (P > 0.05) observed in 0 and 4 d but significant difference(P < 0.05) in left three storage periods between the samples ofcontrol and RF treatment. The increasing content of the fatty acidincreases the occurrence possibility of lipid oxidation reaction andresulted in the increased acidity of the stored corns (Onigbinde andAkinyele, 1988). During the same storage period, the fatty acidcontent in RF treated corns increased at a slower rate than that incontrols, indicating that RF treatments would improve the cornstorage stability.

The initial and final color values of corn samples at 15.0% w.b. inthe polypropylene container before and after RF treatment aresummarized in Table 7. L* values decreased while a* and b* valuesincreased after RF treatment, but there were no significant

Table 6Storage quality characteristics (mean ± SD over 3 replicates) of corn samples with 15.0% w.b. in the polypropylene container before and after radio frequency (RF) treatmentduring the storage.

Quality Treatment Storage time (days) at 35 �C, 68% RH

0 4 8 12 17

Protein (%) Control 5.8 ± 0.2aAa 5.8 ± 0.1aA 5.8 ± 0.05aA 5.7 ± 0.1aA 5.7 ± 0.1aARF 5.8 ± 0.2aA 5.8 ± 0.1aA 5.8 ± 0.1aA 5.8 ± 0.0aA 5.7 ± 0.1aA

Fat (%) Control 5.2 ± 0.0aA 5.2 ± 0.1aAB 5.1 ± 0.09aB 4.9 ± 0.1aBC 4.8 ± 0.0aCRF 5.3 ± 0.0aA 5.2 ± 0.1aAB 5.1 ± 0.08aAB 5.0 ± 0.1aBC 4.9 ± 0.0aC

Starch (%) Control 71.3 ± 0.3aA 71.3 ± 0.2aA 71.0 ± 1.44aA 70.9 ± 0.9aA 69.9 ± 0.7aARF 70.0 ± 0.3aA 71.0 ± 0.4aA 70.9 ± 0.57aA 70.7 ± 0.9aA 70.2 ± 1.0aA

Ash (%) Control 1.2 ± 0.0aA 1.1 ± 0.0aA 1.1 ± 0.02aA 1.1 ± 0.0aA 1.1 ± 0.1aARF 1.1 ± 0.0aA 1.1 ± 0.0aA 1.1 ± 0.01aA 1.1 ± 0.0aA 1.1 ± 0.0aA

Fatty acid (mg/100 g) Control 26.5 ± 0.2aA 30.0 ± 0.1aB 34.9 ± 0.07aC 40.0 ± 0.1aD 48.8 ± 0.1aERF 27.0 ± 0.2aA 30.1 ± 0.1aB 32.4 ± 0.11bC 36.6 ± 0.1bD 42.2 ± 0.1bE

a Different lower and upper case letters indicate that means are significantly different at P ¼ 0.05 among treatments and storage time, respectively.

Table 7Changes in color values of RF treated and untreated corn samples with 15.0% w.b. in the polypropylene container during storage.

Color parameters Storage time (days)b

Treatment 0 4 8 12 17

L*c Control 72.5 ± 4.5a 73.2 ± 4.5 73.7 ± 5.4 74.0 ± 5.1 74.2 ± 4.0RF 72.4 ± 4.4 72.6 ± 3.4 73.0 ± 5.8 73.7 ± 5.2 74.1 ± 4.0

a*c Control (�1.1 ± 1.5) (�1.5 ± 2.7) (�1.9 ± 2.1) (�3.0 ± 1.6) (�4.1 ± 2.0)RF (�0.9 ± 1.4) (�1.2 ± 2.4) (�1.7 ± 2.1) (�2.6 ± 1.6) (�3.5 ± 1.6)

b*c Control 36.2 ± 5.4 35.6 ± 5.7 35.2 ± 4.4 34.6 ± 4.2 34.4 ± 4.0RF 36.3 ± 5.5 36.1 ± 5.8 35.4 ± 4.5 35.0 ± 4.3 34.6 ± 4.0

a No significant difference was observed both for treatments and storage time (P > 0.05).b 17 days at 35 �C, 68% RH to simulate 1-year storage at 10 �C.c L* (lightness); a* (redness-greenness); b* (yellowness-blueness).

A. Zheng et al. / Journal of Stored Products Research 68 (2016) 63e72 71

differences (P > 0.05) in L*, a* and b* values of corn samples be-tween the control and RF treatment. The result was consistent withthat found by Odjo et al. (2012), showing that the temperaturesslightly affect the color values of corn. The L* value of RF treatedsamples slightly increased with increasing storage time, whereas a*and b* values decreased. This trend was likely to be caused by thedecrease of protein content, and is explained in Jamin and Flores(1998) and Sandhu et al. (2007), who found that the a* and b*

values were positively correlated with protein content while the L*

value were negatively correlated with protein content. The resultsshowed that RF treatment protocol did not affect the color stabilityof corn after RF heating and for the accelerated storage time. Thecolor changes of RF treatment corn samples with storage time are inagreement with those of the milled rice (Zhou et al., 2015).

4. Conclusions

RF treatments sharply reduced the heating time and raised theheating rate of corn samples at 15.0% w.b. in the polypropylenecontainer compared to hot air heating, and the optimal heating rateof 6.0 �C/min with the corresponding suitable electrode gap of11.0 cmwas obtained to develop the RF treatment protocol. The RFheating uniformity was improved by adding forced hot air at 70 �C,movement at 6.6 m/h, two mixings, and holding 14 min at the hotair of 70 �C, which were used in the final RF treatment protocol. Forthe corn samples with five moisture levels, the smaller heatinguniformity index appeared in the sample with low MC. The cornsamples at 15.0% w.b. in the polyetherimide container had a betterRF heating uniformity compared to samples in the polystyrene andpolypropylene containers. Corn quality parameters were notaffected by the RF treatments even after the accelerated storage. RFtreatments, therefore, should provide a practical, effective andenvironmentally friendly method for pasteurizing corns while

maintaining product quality. Future research is needed to confirmthe treatment efficacy using inoculated corns and scaled-up forlarge-scale industrial applications.

Acknowledgements

This research was conducted in the College of Mechanical andElectronic Engineering, Northwest A&F University, and supportedby research grants from General Program of National Natural Sci-ence Foundation of China (No. 31371853), Program of IntroducingInternational Advanced Agricultural Science and Technologies (948Program) of Ministry of Agriculture of China (2014-Z21). The au-thors thank to Bo Ling, Zhi Huang, Lixia Hou, and Rui Li for theirhelps in conducting experiments.

References

AOAC, 2000a. AOAC Official Method 945.15: Moisture in Cereal Adjuncts. AOACInternational, Arlington, Washington, DC.

AOAC, 2000b. AOAC Official Method 996.11: Starch in Cereal Products. AOAC In-ternational, Arlington, Washington, DC.

AOAC, 2005. AOAC Official Method 992.23: Crude Protein in Cereal Grains andOilseeds. AOAC International, Arlington, Washington, DC.

AOAC, 2006. AOAC Official Method 992.23: Crude Fat in Feeds and Cereal Grains.AOAC International, Arlington, Washington, DC.

Bankole, S.A., Adenusi, A.A., Lawal, O.S., Adesanya, O.O., 2010. Occurrence of afla-toxin B1 in food products derivable from ‘egusi’ melon seeds consumed insouthwestern Nigeria. Food Control 21, 974e976.

Birla, S., Wang, S., Tang, J., Hallman, G., 2004. Improving heating uniformity of freshfruit in radio frequency treatments for pest control. Postharvest Biol. Technol.33, 205e217.

Briones, V., Aguilera, J.M., 2005. Image analysis of changes in surface color ofchocolate. Food Res. Int. 38, 87e94.

Buzrul, S., Alpas, H., 2007. Modeling inactivation kinetics of food borne pathogens ata constant temperature. LWT-Food Sci. Technol. 40, 632e637.

Fang, Y., Hu, J., Xiong, S., Zhao, S., 2011. Effect of low-dose microwave radiation onAspergillus parasiticus. Food Control 22, 1078e1084.

FAO, 2007. Food and Agriculture Organization. Grain Storage Techniques. Availableat: http://www.fao.org.

A. Zheng et al. / Journal of Stored Products Research 68 (2016) 63e7272

FAOSTAT, 2013. Food and Agriculture Organization of the United Nations StatisticsDivision. Available at: http://www.faostat3.fao.org/browse/Q/*/E.

Gao, M., Tang, J., Villa-Rojas, R., Wang, Y., Wang, S., 2011. Pasteurization processdevelopment for controlling Salmonella in in-shell almonds using radio fre-quency energy. J. Food Eng. 104, 299e306.

Gao, M., Tang, J., Wang, Y., Powers, J., Wang, S., 2010. Almond quality as influencedby radio frequency heat treatments for disinfestation. Postharvest Biol. Technol.58, 225e231.

GB/T 20570-2015, 2015. Guidelines for Evaluation of Maize Storage Character,China, Beijing.

Gu, R., Chen, F., Liu, B., Wang, X., Liu, J., Li, P., Pan, Q., Pace, J., Soomro, A.-A.,Lubberstedt, T., 2015. Comprehensive phenotypic analysis and quantitative traitlocus identification for grain mineral concentration, content, and yield in maize(Zea mays L.). Theor. Appl. Genet. 128, 1777e1789.

Gunterus, A., Roze, L., Beaudry, R., Linz, J., 2007. Ethylene inhibits aflatoxinbiosynthesis in Aspergillus parasiticus grown on peanuts. Food Microbiol. 24,658e663.

Hertwig, C., Reineke, K., Ehlbeck, J., Knorr, D., Schluter, O., 2015. Decontamination ofwhole black pepper using different cold atmospheric pressure plasma appli-cations. Food Control 55, 221e229.

Hou, L., Ling, B., Wang, S., 2014. Development of thermal treatment protocol fordisinfesting chestnuts using radio frequency energy. Postharvest Biol. Technol.98, 65e71.

Huang, Z., Zhang, B., Marra, F., Wang, S., 2016. Computational modelling of theimpact of polystyrene containers on radio frequency heating uniformityimprovement for dried soybeans. Innov. Food Sci. Emerg. Technol. 33, 365e380.

Huang, Z., Zhu, H., Yan, R., Wang, S., 2015. Simulation and prediction of radio fre-quency heating in dried soybeans for thermal disinfestations. Biosys. Eng. 129C,34e47.

Jamin, F.F., Flores, R.A., 1998. Effect of additional separation and grinding on thechemical and physical properties of selected corn dry-milled streams. CerealChem. 75, 166e170.

Jiao, S., Johnson, J., Tang, J., Wang, S., 2012. Industrial-scale radio frequency treat-ments for insect control in lentils. J. Stored Prod. Res. 48, 143e148.

Jiao, Y., Tang, J., Wang, S., 2014. A new strategy to improve heating uniformity of lowmoisture foods in radio frequency treatment for pathogen control. J. Food Eng.141, 128e138.

Jin, Z., Wang, S., Han, P., 2011. Inactivation kinetics model of Aspergillus parasiticus inmouldy maize by microwave processing. Trans. CSAM 42, 148e153.

Kanapitsas, A., Batrinou, A., Aravantinos, A., Markaki, P., 2015. Effect of g-radiationon the production of aflatoxin B1 by Aspergillus parasiticus in raisins (Vitisvinifera L.). Radiat. Phys. Chem. 106, 327e332.

Kim, S.Y., Sagong, H.G., Choi, S.H., Ryu, S., Kang, D.H., 2012. Radio-frequency heatingto inactivate Salmonella Typhimurium and Escherichia coli O157:H7 on black andred pepper spice. Int. J. Food Microbiol. 153, 171e175.

Ling, B., Guo, W., Hou, L., Li, R., Wang, S., 2015. Dielectric properties of pistachiokernels as influenced by frequency, temperature, moisture and salt content.Food Bioprocess Technol. 8, 420e430.

Ling, B., Hou, L., Li, R., Wang, S., 2016. Storage stability of pistachios as influenced byradio frequency treatments for postharvest disinfestations. Innov. Food Sci.Emerg. Technol. 33, 357e364.

Liu, Y., Tang, J., Mao, Z., Mah, J.H., Jiao, S., Wang, S., 2011. Quality and mold control ofenriched white bread by combined radio frequency and hot air treatment.J. Food Eng. 104, 492e498.

Liu, Z., Gao, J., Yu, J., 2006. Aflatoxins in stored maize and rice grains in LiaoningProvince, China. J. Stored Prod. Res. 42, 468e479.

Malumba, P., Janas, S., Masimango, T., Sindic, M., Deroanne, C., Bera, F., 2009. In-fluence of drying temperature on the wet-milling performance and the proteinssolubility indexes of corn kernels. J. Food Eng. 95, 393e399.

Massey, T.E., Stewart, R.K., Daniels, J.M., Liu, L., 1995. Biochemical and molecularaspects of mammalian susceptibility to aflatoxin B1 carcinogenicity. Proceed.Soc. Exp. Biol. Med. 208, 213e227.

Mendez-Albores, A., Cardenas-Rodriguez, D.A., Vazquez-Duran, A., 2014. Efficacy ofmicrowave-heating during alkaline processing of Fumonisin contaminatedmaize. Iran. J. Public Health 43, 147e155.

Odjo, S., Malumba, P., Dossou, J., Janas, S., Bera, F., 2012. Influence of drying andhydrothermal treatment of corn on the denaturation of salt-soluble proteinsand color parameters. J. Food Eng. 109, 561e570.

Onigbinde, A., Akinyele, I., 1988. Biochemical and nutritional changes in corn (Zeamays) during storage at three temperatures. J. Food Sci. 53, 117e117.

Pan, L., Jiao, S., Gautz, L., Tu, K., Wang, S., 2012. Coffee bean heating uniformity and

quality as influenced by radio frequency treatments for postharvest disinfes-tations. Trans. ASABE 55, 2293e2300.

Reddy, K.R.N., Reddy, C.S., Muralidharan, K., 2009. Detection of Aspergillus spp. andaflatoxin B1 in rice in India. Food Microbiol. 26, 27e31.

Reed, C., Doyungan, S., Ioerger, B., Getchell, A., 2007. Response of storage molds todifferent initial moisture contents of maize (corn) stored at 25 C, and effect onrespiration rate and nutrient composition. J. Stored Prod. Res. 43, 443e458.

Sandhu, K.S., Singh, N., Malhi, N.S., 2007. Some properties of corn grains and theirflours I: physicochemical, functional and chapati-making properties of flours.Food Chem. 101, 938e946.

Singh, N., Singh, J., Kaur, L., Sodhi, N.S., Gill, B.S., 2003. Morphological, thermal andrheological properties of starches from different botanical sources. Food Chem.81, 219e231.

Skandamis, P., Nychas, G.J., 2001. Effect of oregano essential oil on microbiologicaland physico-chemical attributes of minced meat stored in air and modifiedatmospheres. J. Appl. Microbiol. 91, 1011e1022.

Sosa-Morales, M., Tiwari, G., Wang, S., Tang, J., Garcia, H., Lopez-Malo, A., 2009.Dielectric heating as a potential post-harvest treatment of disinfestingmangoes, Part II: development of RF-based protocols and quality evaluation oftreated fruits. Biosys. Eng. 103, 287e296.

Tang, J., Ikediala, J.N., Wang, S., Hansen, J.D., Cavalieri, R.P., 2000. High-temperature-short-time thermal quarantine methods. Postharvest Biol. Technol. 21, 129e145.

Taoukis, P.S., Labuza, T.P., Saguy, I.S., 1997. Kinetics of food deterioration and shelf-life prediction. Handb. Food Eng. Pract. 361e403.

Tiwari, G., Wang, S., Birla, S., Tang, J., 2008. Effect of water-assisted radio frequencyheat treatment on the quality of ‘Fuyu’ persimmons. Biosys. Eng. 100, 227e234.

Tiwari, G., Wang, S., Tang, J., Birla, S., 2011. Analysis of radio frequency (RF) powerdistribution in dry food materials. J. Food Eng. 104, 548e556.

Ubwa, S.T., Abah, J., Asemave, K., Shambe, T., 2012. Studies on the gelatinizationtemperature of some cereal starches. Int. J. Chem. 4, 22e28.

Velu, V., Nagender, A., Rao, P.G.P., Rao, D.G., 2006. Dry milling characteristics ofmicrowave dried maize grains (Zea mays L.). J. Food Eng. 74, 30e36.

Wang, K., Zhu, H., Chen, L., Li, W., Wang, S., 2015. Validation of top electrode voltagein free-running oscillator radio frequency systems with different moisturecontent soybeans. Biosys. Eng. 131, 41e48.

Wang, S., Birla, S., Tang, J., Hansen, J.D., 2006a. Postharvest treatment to controlcodling moth in fresh apples using water assisted radio frequency heating.Postharvest Biol. Technol. 40, 89e96.

Wang, S., Yue, J., Chen, B., Tang, J., 2008. Treatment design of radio frequencyheating based on insect control and product quality. Postharvest Biol. Technol.49 (3), 417e423.

Wang, S., Monzon, M., Johnson, J., Mitcham, E., Tang, J., 2007a. Industrial-scale radiofrequency treatments for insect control in walnuts: I: heating uniformity andenergy efficiency. Postharvest Biol. Technol. 45, 240e246.

Wang, S., Monzon, M., Johnson, J., Mitcham, E., Tang, J., 2007b. Industrial-scale radiofrequency treatments for insect control in walnuts: II: insect mortality andproduct quality. Postharvest Biol. Technol. 45, 247e253.

Wang, S., Tang, J., Johnson, J., Mitcham, E., Hansen, J., Cavalieri, R., Bower, J., Biasi, B.,2002. Process protocols based on radio frequency energy to control field andstorage pests in in-shell walnuts. Postharvest Biol. Technol. 26, 265e273.

Wang, S., Tang, J., Johnson, J., Mitcham, E., Hansen, J., Hallman, G., Drake, S.,Wang, Y., 2003. Dielectric properties of fruits and insect pests as related to radiofrequency and microwave treatments. Biosys. Eng. 85, 201e212.

Wang, S., Tang, J., Sun, T., Mitcham, E., Koral, T., Birla, S.L., 2006b. Considerations indesign of commercial radio frequency treatments for postharvest pest controlin in-shell walnuts. J. Food Eng. 77, 304e312.

Wang, S., Tiwari, G., Jiao, S., Johnson, J., Tang, J., 2010. Developing postharvestdisinfestation treatments for legumes using radio frequency energy. Biosys.Eng. 105, 341e349.

Wang, X., Zhang, L., 2011. The impact of quasi-low temperature storage on the cornstorage life. In: Food Storage Technology Innovation and Warehouse Refine-ment Management Seminar, pp. 149e152.

Zhou, L., Ling, B., Zheng, A., Zhang, B., Wang, S., 2015. Developing radio frequencytechnology for postharvest insect control in milled rice. J. Stored Prod. Res. 62,22e31.

Zhou, L., Wang, S., 2016. Verification of radio frequency heating uniformity andSitophilus oryzae control in rough, brown, and milled rice. J. Stored Prod. Res.65, 40e47.

Zia-Ur-Rehman, 2006. Storage effects on nutritional quality of commonly consumedcereals. Food Chem. 95, 53e57.